CMIP6多模式对三江源冬半年极端降水模拟能力评估

Evaluation of CMIP6 Multi-model Simulation Ability of Extreme Precipitation in the Three-rivers in Winter Half Year

为了评估4个全球气候模式(CMIP6)在三江源地区冬半年极端降水的模拟能力,基于台站监测资料和模式历史模拟数据,针对8个极端降水指数,采用泰勒图及多个统计方法进行量化分析。结果表明:(1)多数模式及其集合平均均能够较好地模拟出三江源冬半年极端降水的年际变化趋势,但模式模拟结果均高于观测;(2)模式对8个极端降水指数空间分布模拟能力较强,其空间相关系数较高,大部分达到0.8且均通过0.01显著性;(3)对不同指数而言,模式模拟差异性较大并且模式对于各极端降水指数的时间变率的模拟能力显著弱于空间分布模拟;(4)对不同台站而言,模式对不同台站间的模拟效果差距明显,但总体依旧是多模式集合平均MAM表现最好。

In order to evaluate the simulation ability of four global climate models(CMIP6)for extreme precipitation in the winter half year in the Three-rivers region,eight extreme precipitation indices were quantitatively analyzed by using Taylor plots and multiple statistical methods based on the monitoring data of stations and model historical simulation data.The results show that:(1)Most of the models and their ensemble averages can well simulate the interannual variation trend of extreme precipitation in the winter half year over the Three-rivers,but the model simulation results of the model are higher than the observed ones;(2)The model has a strong ability to simulate the spatial distribution of the eight extreme precipitation indices,and its spatial correlation coefficient is relatively high,most of which reach 0.8 and pass the significance of 0.01.(3)For different indices,the model simulation has a large difference,and its ability to simulate the temporal variability of extreme precipitation indices is significantly weaker than that of spatial distribution simulation.(4)For different stations,there is a significant difference in the simulation effect of the model on different stations,but the average MAM of the multi-mode set is still the best overall.